A physics-based reduced-order model for battery voltage and state of charge prediction
When a high-fidelity electrochemical model of a specific cell is available, a reduced-order formulation can retain its predictive accuracy at a fraction of the computational cost. This work presents a physics-based reduced-order model for battery voltage and state of charge prediction, built on the single-particle model with electrolyte dynamics (SPMe). The reduced-order model is derived from a high-fidelity Doyle–Fuller–Newman (DFN) reference that is parameterised for a large-format NMC pouch cell (57.2 Ah). A quartic polynomial profile approximates the solid-phase diffusion, and a surrogate model represents the electrolyte concentration dynamics through two modes with distinct time scales. The electrolyte surrogate is first fitted to DFN-predicted concentration profiles to establish physically grounded parameter values. These values then serve as regularisation anchors in a joint refinement of the electrolyte and anode solid-diffusion parameters against the terminal voltage. This preserves the identified time-scale structure while improving voltage accuracy. The model is tested against the DFN reference across constant-current and pulse operation, and further compared with experimental measurements on concatenated WLTC and NEDC driving cycles. Under the isothermal conditions studied ( T = 25 ° C ), the reduced-order model achieves voltage and state of charge predictions in close agreement with the DFN reference and experimental data while executing more than three orders of magnitude faster than real time.
Authors
- Martin Kozek (ORCID: https://orcid.org/0000-0003-0402-3309)
- Christoph Hametner (ORCID: https://orcid.org/0000-0003-2912-4771)
- Emiliano Vitaliani
- Stefan Jakubek (ORCID: https://orcid.org/0009-0006-7017-4024)
- Matteas Jelovic (ORCID: https://orcid.org/0009-0003-0777-8640)
- Bernhard Pöllabauer
- Chao Chen
Institutions
- TU Wien (AT)
- Anstalt für Verbrennungskraftmaschinen List (Austria) (AT)
Publication Details
- Journal
- Journal of Energy Storage
- Published
- 2026-11-21
- DOI
- https://doi.org/10.1016/j.est.2026.125768
- Primary Topic
- Battery Modeling and Management
- Type
- article
- Field-Weighted Citation Impact
- 0.00